Application porting optimization on heterogeneous systems

Nikita Storublevtcev, Vladimir Korkhov, Alexey Beloshapko, Alexander Bogdanov

Результат исследований: Научные публикации в периодических изданияхстатьянаучнаярецензирование

2 Цитирования (Scopus)

Выдержка

Modern heterogeneous computer systems offer an exceptional computational potential, but require specific knowledge and experience on the part of the programmer to fully realize it. In this paper we explore different approaches to the task of adapting an application to the heterogeneous computer system. We provide performance evaluation of the test application ported using those approaches. We also evaluate the difficulty and time investment required to implement those approaches in relation to performance improvements they offer.

Язык оригиналаанглийский
Страницы (с-по)25-40
Число страниц16
ЖурналLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Том10963
DOI
СостояниеОпубликовано - 4 июл 2018
Событие18th International Conference on Computational Science and Its Applications, ICCSA 2018 - Melbourne, Австралия
Продолжительность: 2 июл 20185 июл 2018

Ключевые слова

    Предметные области Scopus

    • Теоретические компьютерные науки
    • Компьютерные науки (все)

    Цитировать

    @article{5452eb4fcd5543bf9a6e3a91a94cffbd,
    title = "Application porting optimization on heterogeneous systems",
    abstract = "Modern heterogeneous computer systems offer an exceptional computational potential, but require specific knowledge and experience on the part of the programmer to fully realize it. In this paper we explore different approaches to the task of adapting an application to the heterogeneous computer system. We provide performance evaluation of the test application ported using those approaches. We also evaluate the difficulty and time investment required to implement those approaches in relation to performance improvements they offer.",
    keywords = "Computing, CUDA, Heterogeneous, Performance",
    author = "Nikita Storublevtcev and Vladimir Korkhov and Alexey Beloshapko and Alexander Bogdanov",
    year = "2018",
    month = "7",
    day = "4",
    doi = "10.1007/978-3-319-95171-3_3",
    language = "English",
    volume = "10963",
    pages = "25--40",
    journal = "Lecture Notes in Computer Science",
    issn = "0302-9743",
    publisher = "Springer",

    }

    Application porting optimization on heterogeneous systems. / Storublevtcev, Nikita; Korkhov, Vladimir; Beloshapko, Alexey; Bogdanov, Alexander.

    В: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Том 10963, 04.07.2018, стр. 25-40.

    Результат исследований: Научные публикации в периодических изданияхстатьянаучнаярецензирование

    TY - JOUR

    T1 - Application porting optimization on heterogeneous systems

    AU - Storublevtcev, Nikita

    AU - Korkhov, Vladimir

    AU - Beloshapko, Alexey

    AU - Bogdanov, Alexander

    PY - 2018/7/4

    Y1 - 2018/7/4

    N2 - Modern heterogeneous computer systems offer an exceptional computational potential, but require specific knowledge and experience on the part of the programmer to fully realize it. In this paper we explore different approaches to the task of adapting an application to the heterogeneous computer system. We provide performance evaluation of the test application ported using those approaches. We also evaluate the difficulty and time investment required to implement those approaches in relation to performance improvements they offer.

    AB - Modern heterogeneous computer systems offer an exceptional computational potential, but require specific knowledge and experience on the part of the programmer to fully realize it. In this paper we explore different approaches to the task of adapting an application to the heterogeneous computer system. We provide performance evaluation of the test application ported using those approaches. We also evaluate the difficulty and time investment required to implement those approaches in relation to performance improvements they offer.

    KW - Computing

    KW - CUDA

    KW - Heterogeneous

    KW - Performance

    UR - http://www.scopus.com/inward/record.url?scp=85049948473&partnerID=8YFLogxK

    U2 - 10.1007/978-3-319-95171-3_3

    DO - 10.1007/978-3-319-95171-3_3

    M3 - Article

    VL - 10963

    SP - 25

    EP - 40

    JO - Lecture Notes in Computer Science

    JF - Lecture Notes in Computer Science

    SN - 0302-9743

    ER -